AJP - Lung Information on EB 2010
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Am J Physiol Lung Cell Mol Physiol 294: L942-L954, 2008. First published February 29, 2008; doi:10.1152/ajplung.00240.2007
1040-0605/08 $8.00
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
294/5/L942    most recent
00240.2007v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Web of Science (2)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Brown, S. G.
Right arrow Articles by Wilson, S. M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Brown, S. G.
Right arrow Articles by Wilson, S. M.

The regulation of selective and nonselective Na+ conductances in H441 human airway epithelial cells

Sean G. Brown, Michael Gallacher, Richard E. Olver, and Stuart M. Wilson

Lung Membrane Transport Group, Division of Maternal and Child Health Sciences, College of Medicine, Dentistry and Nursing, Ninewells Hospital and Medical School, University of Dundee, Dundee, Scotland, United Kingdom

Submitted 22 June 2007 ; accepted in final form 25 February 2008


   ABSTRACT
TOP
 ABSTRACT
METHODS
 RESULTS
 DISCUSSION
 GRANTS
 REFERENCES
 
Analysis of membrane currents recorded from hormone-deprived H441 cells showed that the membrane potential (Vm) in single cells (approximately –80 mV) was unaffected by lowering [Na+]o or [Cl–]o, indicating that cellular Na+ and Cl– conductances (GNa and GCl, respectively) are negligible. Although insulin (20 nM, ~24 h) and dexamethasone (0.2 µM, ~24 h) both depolarized Vm by ~20 mV, the response to insulin reflected a rise in GCl mediated via phosphatidylinositol 3-kinase (PI3K) whereas dexamethasone acted by inducing a serum- and glucocorticoid-regulated kinase 1 (SGK1)-dependent rise in GNa. Although insulin stimulation/PI3K-P110{alpha} expression did not directly increase GNa, these maneuvers augmented the dexamethasone-induced conductance. The glucocorticoid/SGK1-induced GNa in single cells discriminated poorly between Na+ and K+ (PNa/PK ~0.6), was insensitive to amiloride (1 mM), but was partially blocked by LaCl3 (La3+; 1 mM, ~80%), pimozide (0.1 mM, ~40%), and dichlorobenzamil (15 µM, ~15%). Cells growing as small groups, on the other hand, expressed an amiloride-sensitive (10 µM), selective GNa that displayed the same pattern of hormonal regulation as the nonselective conductance in single cells. These data therefore 1) confirm that H441 cells can express selective or nonselective GNa (14, 48), 2) show that these conductances are both induced by glucocorticoids/SGK1 and subject to PI3K-dependent regulation, and 3) establish that cell-cell contact is vitally important to the development of Na+ selectivity and amiloride sensitivity.

epithelial Na+ channel; serum- and glucocorticoid regulated kinase 1; phosphatidylinositol 3-kinase; H441 cells; pulmonary Na+ absorption

THE EPITHELIA THAT LINE THE gas exchanging regions of the lung actively absorb water from the overlying film of surface liquid, and this process is driven by electrogenic Na+ transport. This ability to absorb Na+ develops during the very last stages of gestation (10) when it enables the rapid clearance of the liquid that fills the fetal lung lumen. The removal of this liquid is vital if the newborn infant is to breathe at birth, and failure of this absorptive process in premature infants contributes to respiratory distress syndrome (RDS), the most common cause of death among newborn and premature infants in the developed world (38, 39). This capacity to absorb Na+ is retained throughout adult life and is tightly regulated to prevent liquid accumulation in the gas exchanging regions of the lung (54). However, in patients with pulmonary edema, the volume of liquid entering the air spaces overwhelms this absorptive capacity, and, although this life-threatening condition has many possible causes, its eventual resolution is dependent on the capacity of the lungs to absorb Na+ from the air spaces (35, 54). Moreover, in the conducting airways, regulated Na+ transport is an important part of a mechanism that allows precise control over the depth and composition of the overlying liquid film, and dysfunction of this process is believed to play a major role in the development of lung disease in patients with cystic fibrosis, a life-threatening genetic disease (see, for example, Refs. 9, 12, 34).

The integrated functioning of the respiratory tract is therefore dependent on regulated epithelial Na+ transport, and glucocorticoid hormones clearly play an important role in the development and maintenance of this phenotype (see, for example, Refs. 5, 39). Indeed, administration of synthetic glucocorticoids is the most effective therapy for RDS and pulmonary edema. It has, however, been suggested that insulin might also be involved in the control of pulmonary Na+ transport (19), and this hormone does appear to improve lung function in diabetic patients, possibly by stimulating fluid clearance (17, 18). Moreover, insulin clearly regulates Na+ transport in the aldosterone-sensitive epithelia of the distal nephron, and this seems to involve control over the apical abundance and/or activity of the epithelial Na+ channel (ENaC) (see, for example, Refs. 7, 20), a transport protein composed of three homologous subunits ({alpha}-, β-, and {gamma}-ENaC), which appears to allow apical Na+ entry in absorptive epithelia (25). The initial aim of the present study was therefore to determine the extent to which insulin can exert such control over Na+ transport in pulmonary epithelia, and we have therefore 1) recorded membrane currents from a well-characterized, Na+-absorbing cell line derived from the human distal airway epithelium (H441; see, for example, Refs. 14, 33, 48), and 2) used a transient transfection protocol that allowed us to explore the signaling pathways underlying the hormonal control of membrane Na+ conductance (GNa) in these cells.


   METHODS
TOP
 ABSTRACT
 METHODS
RESULTS
 DISCUSSION
 GRANTS
 REFERENCES
 
Cell culture. Standard, serial culture techniques were used to maintain H441 cells in RPMI 1640 culture medium supplemented with FBS (8.5%), newborn calf serum (8.5%), glutamine (2 mM), transferrin (5 µg/ml), selenium (5 ng/ml), and an antibiotic/antimycotic mixture (Sigma, Poole, Dorset, United Kingdom). This differs from the medium used in our earlier studies (11, 14, 24, 45, 56) as it contains no insulin. For experiments, cells removed from culture flasks using trypsin/EDTA were plated onto glass coverslips and incubated (24–48 h) in medium identical to that described above except that the serum components were replaced with FBS (8.5%) that had been dialyzed to remove hormones/growth factors. "Hormone-deprived" cells were maintained under these conditions for 24–48 h before being used in experiments, whereas hormone-treated cells were incubated in medium supplemented with dexamethasone (0.2 µM) or insulin (20 nM) either alone or in combination.

Electrophysiology. Membrane currents (Im) were recorded using the perforated patch recording technique, and only brief details are presented here since this method is described in detail elsewhere (see Refs. 14, 23, 56). The pipette filling solution contained in mM: NaCl, 10; KCl, 18; potassium gluconate, 92; MgCl2, 0.5; EGTA, 1; HEPES, 10; and its pH was adjusted to 7.2 with KOH, which brought K+ to 113.3 mM. The standard bath solution contained in mM: NaCl, 140; KCl, 4.5; MgCl2, 1; CaCl2, 2.5; HEPES, 10; glucose, 5; and its pH was adjusted to 7.4 with NaOH, which brought Na+ to 144.4 mM. Under standard conditions (22°C), the equilibrium potentials for Na+, K+, and Cl– (ENa, EK, and ECl, respectively) were therefore 67.8 mV, –81.9, and –41.9 mV, respectively. The K+-rich (134.5 mM) bath solution was prepared by isoosmotically replacing most Na+ with K+ ([Na+]o = 10 mM), whereas the low Cl– solution ([Cl–]o = 26.5 mM) was prepared by isoosmotically replacing Cl– with either gluconate or I–.

Experimental design and data analysis. Unless otherwise stated, all reported currents were recorded from single cells, although, in some experiments, data were recorded from cells growing in small groups that appeared to contain 3–5 cells. In each experiment, the mean current evoked by driving the holding potential (VHold) through a series of four ramps (87 mV to –113 mV), each of 4-s duration, was recorded, and plots showing the relationship between Im and VHold were constructed. The resting membrane potential (Vm) was inferred from the reversal potential (VRev; i.e., the value of VHold at which Im is 0), whereas total membrane conductance (GTot) was estimated as the slope of the Im-VHold relationship over a physiologically relevant range of potentials (–113 to 0 mV). The input capacitance (Cm) of each preparation was carefully noted, and the magnitudes of all currents and all values of membrane conductance were subsequently normalized to the mean value of Cm associated with a single cell (35 pF). Such data are therefore expressed as picoamperes or nanosiemens per average sized cell (pA per cell or nS per cell). This manipulation was undertaken to ensure that variations between the sizes of different cells did not contribute to the variability in the presented data. The magnitudes of currents recorded from single cells and from small groups of cells can thus be compared directly. Cited values of VHold, Vm, and VRev have all been corrected for the liquid junction potential between the bath and pipette solutions (–13 mV), and, since the bath was always grounded via a salt bridge filled with 3 M KCl/4% agar, the solution changes imposed during the present study had insignificant effects (<1 mV) on this potential (6).

Membrane Na+ current (INa) was quantified by recording Im under control conditions and repeating this measurement 20–30 s after the bath Na+ had been largely ([Na+]o = 10 mM) replaced with N-methyl-D-glucammonium (NMDG+), a nominally impermeant cation. The current that persisted in the presence of NMDG+ was then subtracted from the corresponding record of total current to isolate INa. The N-phenylanthranilic acid (usually known as DPC)-sensitive (IDPC), amiloride-sensitive (IAmil), and LaCl3 (La3+)-sensitive (ILa) components of Im were derived using directly analogous procedures. GNa was estimated by regression analysis of the relationship between INa and VHold, whereas the amiloride-sensitive component of the total membrane conductance (GAmil) was derived by analysis of the relationship between IAmil and VHold. The effects of putative blockers of epithelial Na+ current were quantified by analyzing their effects on the current flowing at –82 ± 5 mV (I–82 mV). This potential was chosen since it equates to EK, which implies that this current can only be carried by Na+ and/or Cl–. I–82 mV was first measured under control conditions (i.e., [Na+]o = 144.4 mM), and the measurement repeated was after [Na+]o had been lowered to 10 mM (NMDG+ substitution). Since this reduction in [Na+]o was imposed with no change in [Cl–]o, the [Na+]o-dependent component of I–82 mV [INa(–82 mV)] gives an estimate of membrane Na+ current. All data are presented as means ± SE, and values of n refer to the number of cells in each group. All reported phenomena were observed in cells from at least three independent passages. The statistical significance of differences between mean values was assessed using Student's unpaired t-test.

Transient transfection protocol. Cells plated onto glass coverslips and maintained in hormone-free medium (see above) for 4–6 h were transfected (Lipofectamine transfection reagent, Invitrogen) with pGL3 plasmids (Invitrogen) incorporating the appropriate cDNA constructs (1 µg; see below) in conjunction with a second plasmid (pEGFP, 0.1 µg) encoding green fluorescent protein (GFP). After 24–48 h of culture under the conditions detailed in the text, coverslips bearing these cells were mounted into a perfusion chamber attached to the stage of a Nikon inverted microscope equipped with epifluorescence optics. Membrane currents were then recorded from successfully transfected cells, which were identified by GFP fluorescence. The success of this method is based on the assumption that the GFP-expressing cells will also express the test constructs. We believe this assumption is justified as the transfection reagent used is designed to form micelle-like structures containing many plasmid DNA molecules. There is no reason why a particular type of plasmid would be excluded from these complexes, and, since the entire micelle is taken up by the transfected cells, there is no mechanism by which a cell could selectively express a GFP-encoding plasmid.

cDNA constructs. Cellular serum- and glucocorticoid-regulated kinase 1 (SGK1; see Ref. 31) activity was artificially increased by transfecting cells with a cDNA construct encoding a glutathione S-transferase (GST)-fusion protein incorporating a truncated form of SGK1 lacking 60 NH2 terminus amino acid residues that had been further modified by mutating Ser422 to Asp (SGK1-S422D). The NH2 terminal truncation induces a 20- to 250-fold increase in protein expression by preventing degradation of the protein, whereas the S422D mutation allows this truncated protein to be more readily activated by 3-phosphoinositide-dependent protein kinase 1 (PDK1) (29, 31). Taken together, these mutations confer a constitutively active phenotype on this mutant (29). Nonspecific effects of the transfection procedure and/or expression of a heterologous protein were controlled for using a construct encoding an analogous GST/truncated SGK1 protein in which Lys127 had been mutated to Ala (SGK1-K127A). This mutation disrupts the protein ATP-binding site, creating a catalytically inactive form of the enzyme (29). The role of phosphatidylinositol 3-kinase (PI3K) was explored using constructs encoding chimeric proteins consisting of the catalytic, P110{alpha} domain of PI3K (PI3K-P110{alpha}) attached to the extracellular and transmembrane domains of the rat CD2 surface antigen, which effectively anchors the PI3K-P110{alpha} subunit to the inner surface of the plasma membrane. The construct used to enhance membrane PI3K activity (rCD2-P110{alpha}) contained wild-type PI3K-P110{alpha}, whereas the corresponding control construct (rCD2-P110{alpha}-R1130P) incorporated a catalytically inactive, mutant form of PI3K-P110{alpha}. This system is detailed elsewhere (47).


   RESULTS
TOP
 ABSTRACT
 METHODS
 RESULTS
DISCUSSION
 GRANTS
 REFERENCES
 
Verification of transient transfection procedure. In an initial series of experiments, membrane currents were recorded from cells that had simply been transfected with GFP-expressing plasmids (see METHODS) to assess the efficacy of the transfection procedure and to determine whether exposure to the transfection reagent and/or expression of a heterologous protein had any effect on the electrical properties of the cells. These experiments showed that 20–40% of the cells expressed sufficient GFP to be visible under epifluorescence. We often saw groups of 3–5 cells that included only a single transfected cell, and, since earlier work had indicated that such cells are electrically interconnected (14), we decided to record currents from single cells to be sure that data were derived from transfected cells. Experiments in which membrane currents were recorded from such GFP-expressing cells (identified by epifluorescence) showed that heterologous expression of this marker protein did not effect the electrical properties of hormone-deprived cells.

Properties of hormone-deprived cells. Figure 1A shows currents recorded from hormone-deprived cells that were initially bathed with the standard, Na+- and Cl–-rich bath solution; these data are derived entirely from single cells to facilitate comparison with data derived from transfected cells (see above). Vm, under these conditions, was approximately –80 mV (Fig. 1B), and this potential did not differ significantly from EK (–81.9 mV). Lowering [Na+]o (10 mM, NMDG+ substitution; Fig. 1A) or [Cl–]o (26.5 mM, gluconate substitution) had no effect on Im and thus caused no change in Vm (Fig. 1B), whereas raising [K+]o to 134.5 mM (Na+ substitution) caused depolarization (Fig. 1B). GK is therefore the dominant ionic conductance under these conditions.


Figure 1
View larger version (14K):
[in this window]
[in a new window]

  		Fig. 1. Electrophysiological properties of hormone-deprived, single cells. A: plot showing the relationship between membrane current (Im) and holding potential (VHold) for cells cultured in nominally hormone-free medium. Currents (n = 5) were recorded during exposure to the standard, Na+-rich bath solution (control) and after this solution had been exchanged for the N-methyl-D-glucammonium (NMDG+)-rich bath solution containing 10 mM Na+ (low Na+). B: values of membrane potential (Vm) derived by analysis of the Im-VHold relationship (see METHODS). Such measurements were first made under control conditions (open columns), and the measurement were repeated after composition of the bath solution had been modified (experimental; filled columns) either by lowering [Na+]o to 10 mM (low Na+; further analysis of data in A), lowering [Cl–]o to 26.5 mM (low Cl–, gluconate substitution, n = 5), or raising [K+]o to 134.5 mM (high K+, n = 8). ***Statistically significant difference from the paired control value (P < 0.001). All data are means ± SE.

 
Effects of insulin. Analysis of currents recorded from insulin-treated (Fig. 2A) cells showed that Vm was normally approximately –60 mV; this potential differed from that measured under hormone-free conditions (see above; P < 0.05) and from EK (P < 0.05), indicating that insulin depolarizes Vm by ~20 mV. Replacing [Na+]o with NMDG+ had no effect on the Im in these cells (Fig. 2A) and so caused no change in Vm (Fig. 2C), whereas lowering [Cl–]o to 26.5 mM (gluconate substitution) led to a rightward shift in the Im-VHold relationship (Fig. 2B). Further analysis showed that this reduction in [Cl–]o depolarized Vm by 16.9 ± 2.9 mV (Fig. 2B).


Figure 2
View larger version (12K):
[in this window]
[in a new window]

  		Fig. 2. Electrophysiological properties of insulin-stimulated single cells. A: the relationship between Im and VHold for cells cultured in insulin-containing medium; currents (n = 5) were recorded both under control conditions and during exposure to the NMDG+-rich, low Na+ solution. B: results of directly analogous experiments that explored the effects of lowering [Cl–]o to 26.5 mM by isoosmotically replacing this anion with gluconate (n = 5). C: values of resting Vm estimated by regression analysis of the data shown in A and B. ***P < 0.01, statistically significant difference from control. All data are means ± SE.

 
CFTR-172 (10 µM, 2–3 min), a relatively selective blocker of the Cl– channels associated with expression of the CFTR protein (36), had no effect (n = 3) on GTot (control: 1.52 ± 0.47 nS per cell; CFTR-172: 1.88 ± 0.5 nS per cell) or Vm (control: –62.7 ± 9.0 mV; CFTR-172: –53.0 ± 9.8 mV) in insulin-treated cells. Subsequent experiments therefore explored the effects of DPC, a nonspecific blocker of epithelial Cl– channels (16). The first such experiments confirmed that Vm in single, hormone-deprived cells was normally approximately –80 mV, and DPC (1 mM) had no effect on Im in such cells (Fig. 3A). Studies of insulin-treated cells (Fig. 3B) confirmed the depolarizing action of this hormone (P < 0.05) and showed that DPC reduced GTot (control: 1.71 ± 0.40 nS per cell; DPC: 1.08 ± 0.3 nS per cell; P < 0.01) and hyperpolarized Vm to a potential that was essentially identical to EK (control: –61.1 ± 4.3 mV; DPC: –86.1 ± 4.3 mV; P < 0.02). DPC thus reverses the insulin-induced depolarization. Analysis of IDPC (see METHODS) showed that this current displayed slight inward rectification and reversed at a potential essentially identical to ECl (–39.8 ± 2.3 mV; Fig. 3C), an observation that confirms that IDPC is Cl– selective. Further analysis of the IDPC-VHold relationship revealed a slope conductance of 0.64 ± 0.13 nS per cell, and, since the corresponding value of GTot was 1.6 ± 0.4 nS per cell, the insulin-induced GCl accounts for ~40% of GTot at physiologically relevant potentials.


Figure 3
View larger version (23K):
[in this window]
[in a new window]

  		Fig. 3. N-phenylanthranilic acid (DPC)-sensitive membrane currents in single cells. In each experiment, membrane currents were recorded from cells bathed with the standard, Na+- and Cl–-rich bath solution under control conditions and after 2–3 min of exposure to 1 mM DPC. A: relationships between Im and VHold for cells maintained in nominally hormone-free medium (n = 4). B: directly analogous data from insulin-treated cells (n = 5) C: the data recorded from insulin-treated cells were further analyzed to isolate IDPC (see METHODS), which is plotted (n = 5) against VHold. D: relationships between Im and VHold for cells (n = 5) maintained in nominally hormone-free medium and transfected with rCD2-P110{alpha}-R1130P. E: analogous data (n = 5) from hormone-deprived cells expressing with rCD2-P110{alpha}. F: the relationship between IDPC and VHold (n = 5) for rCD2-P110{alpha}-expressing cells. All data are means ± SE.

 
Expression of constructs encoding wild-type and mutant PI3K-P110{alpha}. Figure 3D shows currents recorded from hormone-deprived cells expressing rCD2-P110{alpha}-R1130P, which incorporates a catalytically inactive PI3K-P110{alpha} subunit (see METHODS). These currents were essentially identical to control (Fig. 3A), and DPC (1 mM, 1–2 min) had no effect (Fig. 3D) on either GTot (control: 0.92 ± 0.37 nS per cell; DPC: 0.94 ± 0.40 nS per cell) or Vm (control: –84.9 ± 1.6 mV; DPC: –79.6 ± 3.2 mV). Expression of this control construct therefore has no effect on the conductive properties of hormone-deprived cells. Expression of rCD2-P110{alpha}, which incorporates wild-type PI3K-P110{alpha} (47), depolarized Vm (P < 0.02) to approximately –60 mV (Fig. 3E) and caused a rise in GTot when compared with data from rCD2-P110{alpha}-R1130P-expressing cells (P < 0.05). Application of DPC inhibited GTot (control: 2.3 ± 0.66 nS per cell; DPC: 1.1 ± 0.33 nS per cell; P < 0.05) and hyperpolarized Vm to a value close to EK (control: –64.0 ± 3.7 mV; DPC: –85.1 ± 3.1 mV; P < 0.001). Further analysis showed that expression of rCD2-P110{alpha} induced a DPC-sensitive current that was two to threefold larger than that seen in insulin-stimulated cells but otherwise essentially identical (Fig. 3F).

Properties of the insulin-induced Cl– conductance. Lowering [Cl–]o to 26.5 mM by isoosmotically substituting I– hyperpolarized insulin-treated cells by ~12 mV (control: –62.6 ± 5.3 mV; I–: –75.9 ± 3.7 mV; n = 4; P < 0.05). Although this result suggests that the insulin-induced GCl may be more permeable to I– than to Cl– (see also Ref. 56), it is also possible that I– may block GCl, which would allow Vm to move toward EK. We therefore undertook further studies of insulin-treated cells that were initially bathed with K+- and Cl–-rich solution containing bupivacaine (3 mM), a local anesthetic that causes essentially complete block of the channels that underlie GK in resting H441 cells (24). The currents recorded under these conditions normally (i.e., when [Cl–]o was 151.5 mM) reversed at a potential close to ECl (–39.7 ± 2.1 mV), and lowering [Cl–]o to 26.5 mM by isoosmotically substituting a nominally impermeant anion (gluconate) reduced the outward current flowing at positive values of VHold and depolarized VRev (P < 0.05) to –17.0 ± 5.0 mV (Fig. 4). These findings (Fig. 4) show that the current recorded under these conditions is primarily carried by anions. Replacing Cl– with I– ([Cl–]o = 26.5 mM) enhanced the outward currents flowing at positive values of VHold and hyperpolarized VRev to –51.3 ± 3.5 mV (P < 0.02; Fig. 4). Since EK was held at ~0 mV throughout all solution changes (i.e., [K+]o was always elevated), this effect shows that exposure to the I–-rich saline must augment a hyperpolarizing anion current. The channels that underlie GCl must therefore be more permeable to I– than to Cl– (see also Ref. 56).


Figure 4
View larger version (11K):
[in this window]
[in a new window]

  		Fig. 4. Bupivacaine-resistant membrane currents in single, insulin-treated cells. All solutions used in these experiments contained bupivacaine (3 mM), which was included to block GK (24). The figure shows relationships between bupivacaine-resistant component of Im and VHold. In each experiment, data were initially derived from cells bathed with the K+ (134.5 mM)- and Cl– (151.5 mM)-rich bath solution, which was designed to hold EK at ~0 mV, and the measurements were then repeated after [Cl–]o had been lowered to 26.5 mM by isoosmotically replacing this anion with gluconate (Gluc.) or I–. All data are means ± SE.

 
Effects of dexamethasone. Analysis of data recorded from dexamethasone-stimulated cells (Fig. 5A) showed that this synthetic glucocorticoid depolarized (P < 0.05) Vm to a value (approximately –60 mV) that differed significantly (P < 0.05) from EK. Lowering [Cl–]o (gluconate substitution) had no effect on Vm in dexamethasone-stimulated cells (Fig. 5B), whereas reducing [Na+]o inhibited the inward currents at negative values of VHold (Fig. 5A) and hyperpolarized Vm (Fig. 5B). The dexamethasone-induced depolarization thus reflects a rise in GNa (see also Ref. 14) with no change in GCl. Figure 6A shows values of INa derived from these data (see METHODS) and plotted against VHold and also includes equivalent data from hormone-deprived cells (Fig. 1A). It is abundantly clear that INa is negligible in hormone-deprived cells and that dexamethasone induces a clear current. Regression analysis indicated that this synthetic glucocorticoid increased GNa to ~0.6 nS per cell (Fig. 6C). Figure 6B shows equivalent data from cells stimulated with insulin either alone or in combination with dexamethasone. These data confirm that insulin does not directly induce INa but establish that this hormone can augment the dexamethasone-induced GNa (Fig. 6, B and C).


Figure 5
View larger version (12K):
[in this window]
[in a new window]

  		Fig. 5. Electrophysiological properties of single, dexamethasone-stimulated cells. A: plot showing the relationship between Im and VHold for cells (n = 9) maintained in dexamethasone-containing (0.2 µM) medium for ~24 h; currents were recorded both under control conditions and after [Na+]o had been lowered to 10 mM (NMDG+ substitution). B: values of Vm estimated from data recorded under conditions (open columns) and then after the ionic composition of the bath solution had been modified (experimental; filled columns) by lowering [Na+]o to 10 mM (low Na+; further analysis of data in A) or lowering [Cl–]o to 26.5 mM (low Cl–, n = 4). All data are means ± SE. ***P < 0.01, statistically significant difference from control.

 

Figure 6
View larger version (21K):
[in this window]
[in a new window]

  		Fig. 6. Hormonal and serum- and glucocorticoid-regulated kinase 1 (SGK1)-induced control of membrane Na+ current (INa) in single cells. A: plots showing the relationship between INa and VHold for cells cultured in nominally hormone-free medium (n = 5) or exposed to dexamethasone (Dexameth., Dex.) for ~24 h (n = 9). B: equivalent data from cells exposed to insulin for ~24 h either alone (n = 13) or in combination with dexamethasone (n = 9). C: values of cellular Na+ conductance (GNa) derived by analysis of the data in A and B (see METHODS); **P < 0.02, ***P < 0.01, statistically significant difference from control; {dagger}statistically significant effect of insulin. D: relationship between INa and VHold for hormone-deprived cells expressing SGK1-K127A (n = 5) or SGK1-S422D (n = 8). E: equivalent data from insulin-treated cells expressing SGK1-K127A (n = 10) or SGK1-S422D (n = 11). F: values of GNa derived from analysis of the data in D and E; *P < 0.05, ***P < 0.01, statistically significant differences between SGK1-K127A and SGK1-S422D; {dagger}statistically significant effects of insulin. All data are means ± SE.

 
INa in cells expressing mutant forms of SGK1. Figure 6D shows the relationships between INa and VHold for hormone-deprived cells expressing SGK1-K127A or SGK1-S422D. INa is negligible in the SGK1-K127A-expressing cells, confirming that the transfection procedure has no effect on membrane conductance. Expression of SGK1-S422D, on the other hand, induced an inward current similar to that seen in dexamethasone-stimulated cells (Fig. 6, D and F). Figure 6E shows analogous data derived from insulin-stimulated cells. Once again, SGK1-K127A expression did not induce any discernible INa, whereas SGK1-S422D expression evoked a substantial current (Fig. 6E). Although insulin did not induce INa in SGK1-K127A-expressing cells (Fig. 6, B and C), this hormone augmented the current evoked by SGK1-S422D (Fig. 6, E and F).

INa in cells expressing rCD2-P110{alpha} and rCD2-P110{alpha}-R1130P. Figure 7A shows data from rCD2-P110{alpha}-R1130P-expressing cells that had been maintained (~24 h) in hormone-free or dexamethasone-supplemented medium. INa was negligible under hormone-free conditions, and dexamethasone induced a current very similar to that seen in untransfected cells (Fig. 6A). Expression of rCD2-P110{alpha}-R1130P therefore has no effect on the conductive properties of unstimulated or dexamethasone-stimulated cells. Figure 7B shows equivalent data from rCD2-P110{alpha}-expressing cells. Although the expression of this construct, which incorporates wild-type PI3K-P110{alpha} (47), did not induce INa in unstimulated cells, it augmented the dexamethasone-induced INa (Fig. 7C).


Figure 7
View larger version (11K):
[in this window]
[in a new window]

  		Fig. 7. Control of INa by phosphatidylinositol 3-kinase (PI3K) in single cells. A: plots showing the INa-VHold relationship for cells expressing the control, rCD2-P110{alpha}-R1130P construct were either maintained in hormone-free medium (n = 5) or exposed to 0.2 µM dexamethasone for 24 h (n = 9). B: plots showing the INa-VHold relationship for hormone-deprived (n = 5) or dexamethasone-stimulated (n = 6) cells expressing rCD2-P110{alpha}. C: values of GNa derived by analysis of the data in A and B; *P < 0.05, ***P < 0.01, statistically significant effects of dexamethasone; {dagger}statistically significant (P < 0.05) difference between the dexamethasone-treated cells expressing rCD2-P110{alpha}-R1130P and rCD2-P110{alpha}. All data are means ± SE.

 
Properties of the hormonally induced GNa. Figure 8A shows the effects of amiloride (10 µM) on the current recorded from single hormone-treated (0.2 µM dexamethasone, 20 nM insulin, ~24 h) cells. The mean value of Cm associated with these recordings was thus ~30 pF. Amiloride has no effect on the recorded current and therefore had no effect on Vm (control: –27.6 ± 2.0 mV; amiloride: –28.5 ± 2.2 mV) or GTot (control: 2.0 ± 0.6 nS per cell; amiloride: 2.1 ± 0.6 nS per cell). Subsequently, reducing [Na+]o to 10 mM (NMDG+ substitution), in the continued presence of amiloride, reduced the inward currents flowing at negative values of VHold and hyperpolarized Vm to approximately –50 mV (–48.3 ± 3.0 mV; P < 0.002). Further analysis of these data confirmed that IAmil was negligible but verified that hormone-stimulated cells expressed a clear INa (Fig. 8B). Further studies of such cells showed that amiloride had no effect on INa when used at 1 mM (n = 5). The hormonally induced INa in single cells is therefore resistant to the effects of amiloride.


Figure 8
View larger version (13K):
[in this window]
[in a new window]

  		Fig. 8. Properties of GNa in single, hormone-treated cells. All data are derived from single cells maintained (~24 h) in medium containing dexamethasone (0.2 µM) and insulin (20 nM), and the value of the input capacitance (Cm) associated with these recordings was 32 ± 1.9 pF. A: in each experiment, currents were recorded under standard conditions (control), during exposure to standard bath solution containing 10 µM amiloride (Amil.), and during exposure to NMDG+-rich, low Na+ solution containing 10 µM amiloride (low Na+). B: these data were further analyzed to isolate IAmil and INa (see METHODS), and these data are plotted against VHold. All data are means ± SE.

 
The data in Fig. 9A were recorded from hormone-treated cells at identical passage to those that provided the data in Fig. 8, and the only difference between these experiments is that all data in Fig. 9 are derived from cells growing in groups that appeared to contain 3–5 cells. The value of Cm associated with these recordings was essentially identical that reported in a previous study of H441 cells (~70 pF; see Ref. 14). Analysis of the present data showed that Vm was normally –3.7 ± 3.8 mV, although, as reported previously (14), there was variability between the values of Vm measured in different cells (range: –20.7 to 20.2 mV), and analysis of the data from all experiments included in the present study in which membrane currents were recorded from such groups of cells showed that the mean value of Vm was approximately –20 mV. This is very similar to the previously reported value (14). Under these conditions, amiloride (10 µM, 30–45 s) hyperpolarized Vm to –56.7 ± 1.5 mV (P < 0.005) and reduced the inward currents flowing at negative values of VHold. Examination of the IAmil-VHold relationship showed that the value of VRev associated with this current was essentially identical to ENa (Fig. 9B), indicating that IAmil is Na+ selective (see also Ref. 14). Subsequently, lowering [Na+]o to 10 mM (NMDG+ substitution), in the continued presence of 10 µM amiloride, caused further hyperpolarization to –68.1 ± 2.4 mV (P < 0.05), which also accords with the earlier data (14), and analysis of INa(–82 mV) (see METHODS) indicated that amiloride caused 92.8% ± 1.2% inhibition (P < 0.005) of INa under these conditions (Fig. 9B).


Figure 9
View larger version (17K):
[in this window]
[in a new window]

  		Fig. 9. Properties of the hormonally induced GNa in cells growing as small groups. A: plots showing the relationship between Im and VHold derived from experiments (n = 5) in which Im was recorded from groups of 3–5 cells (Cm = 71.6 ± 6.3 pF) that had been incubated (~24 h) in medium containing dexamethasone (0.2 µM) and insulin (20 nM). In each experiment, Im was initially recorded under standard conditions (control), and the measurements were repeated during exposure to 10 µM amiloride and the NMDG+-rich, low Na+ solution containing 10 µM amiloride (low Na+). B: these data were further analyzed to isolate IAmil and INa, which are plotted (means ± SE) against VHold. C: plots showing the relationship between IAmil and VHold derived for cells growing as small groups that had been exposed to 0.2 µM dexamethasone (~24 h) either alone (n = 6, Cm = 57.7 ± 3.1 pF) or in combination with 20 nM insulin (n = 10, Cm = 74.1 ± 5.6 pF). D: values of total membrane conductance (GAmil) derived from further analysis of these data (see METHODS) and presented as means ± SE. *P < 0.05, statistically significant differences between the data derived from the 2 groups of cells.

 
The data in Fig. 9C are derived from further studies of cells growing in small groups. These cells had been maintained (~24 h) in medium containing dexamethasone (0.2 µM), either alone or in combination with insulin (20 nM). IAmil was clearly measurable in all cells, and this current consistently reversed at a potential close to ENa. These data thus confirm that glucocorticoids induce a selective GNa in such cells and show that insulin does not influence the ionic selectivity of the underlying conductance. However, it was also clear that insulin could augment this dexamethasone-induced conductance (Fig. 9D).

Pharmacological properties of the hormonally induced conductance in single cells. Further studies confirmed that 1 mM amiloride had no effect on the currents recorded from single, hormone-treated (~24 h, 0.2 µM dexamethasone, 20 nM insulin) cells (Fig. 10A; control Vm: –29.0 ± 9.2 mV; amiloride Vm: –29.4 ± 5.4 mV); further analysis of these data therefore confirmed that IAmil was negligible (Fig. 10B) and revealed no significant inhibition of INa(–82 mV) (Fig. 10C). However, subsequent application of La3+ (1 mM), in the continued presence of amiloride, rapidly (2–5 s) inhibited the inward current flowing at negative values of VHold and hyperpolarized (P < 0.05) Vm to –45.0 ± 8.4 mV (Fig. 10A), indicating that this cation blocks depolarizing Na+ entry. However, lowering [Na+]o to 10 mM (NMDG+ substitution), in the continued presence of La3+, caused further inhibition of the inward current (Fig. 10A) and additional hyperpolarization (–72.6 ± 7.2 mV; P < 0.05), indicating that this block is incomplete. Further analysis showed that La3+ caused ~70% inhibition of INa(–82 mV) (Fig. 10C), although there was some variability between cells since essentially complete block was observed in two instances. This protocol was also used to explore the effects of 15 µM dichlorobenzamil (DCB) and 100 µM pimozide, both of which have been reported to block Na+-permeable channels in other cell types. In each such experiment, the cells were initially exposed to 1 mM amiloride, and these data consistently confirmed that this ENaC blocker had no effect on Im. However, DCB caused ~15% inhibition of INa(–82 mV), whereas pimozide reduced this current by ~45% (Fig. 10D).


Figure 10
View larger version (19K):
[in this window]
[in a new window]

  		Fig. 10. Properties of the hormonally induced (0.2 µM dexamethasone, 20 nM insulin, ~24 h) GNa in single cells. A: in each experiment, Im was initially recorded from cells bathed with the standard bath solution (control), and the measurements were repeated during exposure to standard bath solution containing 1 mM amiloride, after 1 mM LaCl3 (La3+) had also been added to this solution, and finally after this solution had been replaced with NMDG+-rich, low Na+ (10 m) saline containing 1 mM amiloride and 1 mM LaCl3 (low Na+). B: further analysis of the data in A showing the relationships between VHold, IAmil, and ILa (means ± SE). C: the effects of amiloride and La3+ on INa(–82 mV) (see METHODS) were quantified (% inhibition) and presented (means ± SE) together with the results of separate experiments that used this protocol to quantify the effects of 15 µM dichlorobenzamil (DCB; n = 5) and 100 µM pimozide (Pim.; n = 6); ***P < 0.001, statistically significant inhibitory actions, Student's paired t-test. D: plots showing the relationship between ILa and VHold derived from cells bathed with either the standard bath solution (Na+; n = 7) or K+-rich solution containing only 10 mM Na+ (K+; n = 9). These solutions both contained bupivacaine to block the selective K+ conductance in these cells (24).

 
Ionic selectivity of the La3+-sensitive conductance. As well as blocking inward Na+ current, La3+ reduced the outward current flowing at positive values of VHold (Fig. 10A), and analysis of the ILa-VHold relationship (Fig. 10C) indicated that this current reversed at –10.8 ± 6.6 mV. This potential differed from ENa, suggesting that ILa is not Na+-selective. This situation was further analyzed by assigning the K+ permeability (PK) of the underlying conductance a value of unity and reiteratively adjusting the Na+ permeability (PNa) to identify a solution to the Goldman-Hodgkin-Katz (GHK) Equation consistent with the observed value of VRev. This analysis indicated that PNa/PK was 0.50. However, we were concerned that the La3+-induced inhibition of outward current (Fig. 10A) might be due, at least in part, to block of the selective K+ conductance (GK) in these cells (24) and therefore undertook further experiments in which ILa was quantified in the presence of bupivacaine (3 mM), a substance that causes essentially complete block of GK (24). This substance had no effect on ILa (Fig. 10D), and the value of VRev derived from these experiments (–3.8 ± 4.4 mV) did not differ significantly from that reported above. Analysis of these data indicated that PNa/PK was 0.69. Moreover, raising [K+]o to 134.5 mM (Na+ substitution) caused a rightward shift in the ILa-VHold relationship (Fig. 10D) so that VRev was now 6.7 ± 3.8 mV, and this potential did not differ significantly from that predicted by the GHK Equation (4.1 mV) for a conductance with this degree of Na+ selectivity.

Characteristics of the SGK1-S422D induced GNa. Further studies (n = 5) of SGK1-S422D-expressing cells maintained (~24 h) in insulin-containing (20 nM) medium showed that amiloride (10 µM) had no effect on conductive properties of the plasma membrane (control Vm: –22.7 ± 4.5 mV; amiloride Vm: –21.6 ± 3.5 mV), whereas reducing [Na+]o to 10 mM (NMDG+ substitution) consistently hyperpolarized Vm to –42.7 ± 3.9 mV (P < 0.05). Amiloride also had no effect at 1 mM (n = 5). These studies therefore confirm that heterologous expression of SGK1-S422D induces GNa and show that this conductance is amiloride resistant. Further experiments showed that this SGK1-induced conductance was blocked by La3+ (~95% inhibition), pimozide (~50% inhibition), and DCB (~15% inhibition) (Fig. 11A), and, although this pattern is broadly similar to that seen in hormonally stimulated cells, La3+ acted as a slightly more effective blocker of GNa in the SGK1-expressing cells (P < 0.005), causing essentially complete block in most experiments. Experiments in which ILa was quantified during exposure to bupivacaine (3 mM) showed that this current normally reversed at –12.8 ± 1.7 mV, and, although slightly more negative than the equivalent data derived from hormone-treated cells (Fig. 10D), this difference was not statistically significant, and analysis using the GHK Equation (see above) showed that PNa/PK was 0.46. Exposing these cells to the K+-rich bath solution depolarized VRev to 4.8 ± 4.5 mV (Fig. 11B), and this potential is essentially identical to that predicted by the GHK Equation (4.2 mV) for a conductance displaying this degree of Na+ selectivity.


Figure 11
View larger version (14K):
[in this window]
[in a new window]

  		Fig. 11. Properties of the SGK1-S422D-induced GNa in single cells. Membrane currents were recorded from SGK1-S422D-expressing cells maintained (~24 h) in medium supplemented with 20 nM insulin; experiments were undertaken, and data were analyzed using the protocols shown in Fig. 10. A: inhibition of INa (%, means ± SE) by amiloride (1 mM, n = 5), DCB (15 µM, n = 8), pimozide (100 µM, n = 5), and La3+ (1 mM, n = 5); asterisks denote statistically significant inhibition (Student's paired t-test). B: plots showing the relationship (means ± SE) between ILa and VHold for cells bathed with either the standard (Na+, n = 5) or K+-rich bath solution containing 10 mM Na+ (n = 5); both solutions contained 3 mM bupivacaine.

 

   DISCUSSION
TOP
 ABSTRACT
 METHODS
 RESULTS
 DISCUSSION
GRANTS
 REFERENCES
 
The insulin-induced depolarization. Studies of hormone-deprived cells showed that Vm was essentially identical to EK and demonstrated that this potential was unaffected by large reductions in [Na+]o or [Cl–]o, whereas increasing [K+]o caused depolarization. Taken together, these observations show that GNa and GCl are both negligible under these conditions and demonstrate that the electrical properties of hormone-deprived cells are dominated by GK. However, the present estimate of Vm was more negative than the value reported in a previous study in which H441 cells were maintained in glucocorticoid-free medium (14), and it is therefore interesting that insulin was routinely added to all culture media used in this earlier study whereas this hormone was omitted from the media used here (see METHODS). This raised the possibility that insulin might depolarize the cells, and this hypothesis was confirmed by experiments that showed ~24 h exposure to 20 nM insulin depolarized Vm by ~20 mV. However, although insulin has been shown to stimulate Na+ absorption in pulmonary (19) and renal (see, for example, Ref. 7) epithelia, lowering [Na+]o did not hyperpolarize insulin-stimulated cells, whereas replacing [Cl–]o with gluconate caused further depolarization. The insulin-induced depolarization therefore cannot be attributed to an increase in GNa and seems to reflect a rise in GCl that would allow Vm to move away from EK. Subsequent studies confirmed this hypothesis by identifying an insulin-induced, DPC-sensitive current that reversed at a potential close to ECl and accounted for ~40% of GTot at physiologically relevant values of VHold.

As far we are aware, these data provide the first description of insulin-induced GCl in absorptive epithelia, although this hormone has been shown to increase GCl in a liver-derived cell line, apparently by activating PI3K (28), an enzyme that catalyzes the phosphorylation of phosphatidylinositol 3-phosphate and phosphatidylinositol 3,4-bisphosphate to phosphatidylinositol 4,5-bisphosphate (PIP2) and phosphatidylinositol 3,4,5-trisphosphate (PIP3), respectively (see, for example, Ref. 15). These anionic phospholipids control many aspects of cell physiology by activating PI3K and downstream kinases such as SGK1 (15, 29) and by directly controlling ion channel activity (see, for example, Refs. 20, 22). Since PI3K has also been reported to control GCl in arterial smooth muscle (13), subsequent experiments explored the effects of artificially increasing cellular PI3K activity by transiently expressing membrane-anchored PI3K-P110{alpha} (see Ref. 47). Initial studies showed that mock transfection (i.e., heterologous expression of GFP) had no effect on the conductive properties of the cells and established that expression of a control construct encoding a catalytically inactive, membrane-anchored PI3K-P110{alpha} mutant was similarly ineffective. It is therefore clear that the transfection procedure itself has no direct effect on the conductive properties of H441 cells. However, expression of the construct encoding membrane-anchored wild-type PI3K-P110{alpha} depolarized the cells by inducing a DPC-sensitive current similar to that seen in insulin-stimulated cells. This occurred with no effect on GNa (see below), and so artificially augmenting PI3K activity mimics the depolarizing action of insulin.

The insulin/PI3K-induced conductance. We initially sought to characterize the insulin/PI3K-induced GCl by quantifying IDPC in the presence of different external anions, but this proved unfeasible as anionic substitution/DPC application frequently caused rupture of the seal between the recording pipette and the cell, making it essentially impossible to record Im throughout several such solution changes. We therefore adopted an alternative strategy based on the fact that bupivacaine blocks the channels that underlie GK in H441 cells (24). Initial studies of bupivacaine-treated cells bathed with K+- and Cl–-rich saline showed that Vm normally lay close to ECl. Moreover, lowering [Cl–]o (gluconate substitution) caused depolarization, and these findings confirm that GCl is the dominant conductance under these conditions (24). The current recorded under these conditions displayed slight outward rectification, which was surprising since IDPC was inwardly rectifying. The reason for this discrepancy is unknown, but it may be relevant that DPC is a relatively weak anion channel blocker and that efficacy of blockade is influenced by Vm (see, for example, Ref. 44). These effects could give rise to disparities between IDPC and the true Cl– current. Although we cannot formally exclude the possibility that bupivacaine may directly modify the conductive properties of the anion channels that underlie GCl, we are not aware of any precedents for such behavior. However, the most important point to emerge from these studies was that the population of anion channels that underlies GCl was more permeable to I– than to Cl–. The insulin/PI3K-induced GCl therefore resembles an anion-selective conductance recently identified in rat cortical lens fiber cells (55). Although it has been suggested that H441 cells express significant amounts of CFTR (30, 33), these channels are characteristically less permeable to I– than to Cl–, and so the present data, in common with earlier work (56), suggest that CFTR does not contribute significantly to GCl under the present conditions. Further evidence of this came from that fact that CFTR-172, which blocks CFTR with a reasonable degree of selectivity (36), had no effect on the conductive properties of insulin-treated cells.

The dexamethasone-induced depolarization. Dexamethasone also depolarized H441 cells, but this response, in contrast to the response to insulin, involved a rise in GNa (see also Ref. 14) that occurred with no change in GCl. Moreover, this effect was reproduced by transient expression of constitutively active SGK1 (SGK1-S422D), whereas the inactive, control construct (SGK1-K127A) had no effect. These observations are consistent with a body of data that highlights SGK1 as an important regulator of GNa in absorptive epithelia (1–3, 31, 49–51, 53, 57). Although it is clear that glucocorticoid hormones do induce SGK1 expression (31), SGK1 is an important downstream target of PI3K (29), and data from several absorptive cell types have suggested that insulin can increase the apical abundance of ENaC via a mechanism dependent on a PI3K-mediated increase in SGK1 activity (2, 8, 46, 51). Moreover, PIP2 and PIP3 also activate ENaC by binding to the channel complex, and this provides another mechanism by which PI3K-coupled agonists, such as insulin, can stimulate Na+ transport (20, 42, 43, 52). However, although the present data show that insulin stimulation/PI3K-P110{alpha} expression can control GCl, these maneuvers had no effect on GNa in hormone-deprived cells but augmented the dexamethasone/SGK1-induced GNa. The present data indicate that PI3K can only control GNa in cells that have been "primed" by glucocorticoid stimulation/SGK1-S422D expression.

This was surprising since previous studies of H441 cells indicated that PI3K played a central role in the control of Na+ transport (51). However, the cells used in this earlier study displayed a Na+-absorbing phenotype under basal conditions (51), whereas, in our hands, GNa is negligible in hormone-deprived cells (see also Ref. 14). This almost certainly reflects differences in culture conditions since the earlier data (51) are derived from cells maintained in media supplemented with serum that contains a complex mixture of hormones and growth factors that can activate SGK1 (31). Serum stimulation might therefore "prime" H441 cells and thus allow PI3K to contribute control GNa. However, some renal cell lines clearly display basal Na+ transport when deprived of serum/hormones, and it is clear that insulin/PI3K can regulate Na+ transport in these cells (2, 7, 8, 37, 40), indicating that the priming effect described here is not necessary in all experimental systems. Moreover, in renal cells, insulin has no effect on the Na+ current induced by constitutively active SGK1 (2, 4), whereas the SGK1-S422D-induced Na+ current reported here was clearly enhanced by insulin. The physiological significance of these discrepancies is unknown, but they could reflect differences in culture conditions or be related to tissue-related differences in basal activity of regulatory enzymes such as PI3K/SGK1. It would therefore be interesting to record membrane currents from renal cells maintained under conditions absolutely identical to those used here.

Biophysical properties of the hormonally induced conductance. Although the hormonally induced conductance in single cells was permeable to Na+, it was unaffected by high concentrations of amiloride but could be blocked by La3+ and, to a lesser extent, by pimozide and DCB. Biophysical analysis showed that the ion channels underlying this conductance discriminated between Na+ and K+ very poorly (PNa/PK ~0.6), and so the dexamethasone-induced GNa in single cells reflects the expression of a nonselective cation conductance that is subject to further upregulation by insulin. Moreover, transient expression of constitutively active SGK1 mimicked this effect of dexamethasone, and, as far as we are aware, this observation provides the first indication that this important regulatory enzyme can control the activity of such cation channels.

We were, however, extremely surprised by these findings since Na+ entry in absorptive epithelial cells, such as H441 cells, is almost invariably attributed to the Na+-selective channels associated with {alpha}-, β-, and {gamma}-ENaC coexpression, and these are characteristically sensitive to amiloride. Indeed, previous work from this laboratory (14) has shown that glucocorticoids can induce such an ENaC-like GNa in H441 cells. However, almost all data in this earlier report were derived from groups of two to five electrically coupled cells, whereas membrane currents were routinely recorded from single cells in the present study. The reason for this discrepancy is that the present study aimed to explore the physiological consequences of expressing the SGK1/PI3K-encoding constructs, and, since the transfection efficiency was low, we almost never observed cell clumps in which every cell was transfected. Currents were therefore recorded from single cells to ensure that all data originated from unambiguously transfected cells, and the studies of hormonally stimulated cells followed the same protocol in the interests of consistency. However, this clear discrepancy with earlier work prompted us to undertake further studies of cells growing as small groups, and these experiments clearly confirmed (14) that the glucocorticoid-induced GNa recorded under these conditions is Na+ selective and amiloride sensitive. Moreover, the new data presented here extend on these earlier findings by showing this dexamethasone-induced conductance is further upregulated by insulin, which supports the view that this hormone might play a role in the control of pulmonary Na+ transport (17–19).

It is therefore clear that hormonally stimulated H441 cells can express either an amiloride-sensitive, Na+-selective conductance (14) or a La3+-sensitive, nonselective cation conductance, and it is therefore interesting that Jain and colleagues (27) also identified such conductances in alveolar type II cells. Moreover, in this study, the relative abundance of these two channel types was strongly influenced by the culture conditions. Nonselective channels were therefore the dominant conductance in cells that had been cultured on impermeable supports using standard techniques, whereas selective ENaC-like Na+ channels were the main channel type found in cells maintained in steroid-supplemented media and grown to confluence on permeable supports at an air-liquid interface (27). Similarly, studies of the Na+ channels in A549 alveolar cells suggested that dexamethasone stimulation caused an increase in Na+ selectivity that was associated with a reduction in single channel conductance (32). These earlier studies (27, 32) therefore suggest that hormonal stimulation, the nature of the growth substrate, the composition of the culture medium, and the methods used to culture the cells can all influence the relative expression of the selective and nonselective channels. Moreover, assays of membrane protein abundance and experiments using anti-sense oligonucleotides to suppress expression of individual ENaC subunits indicated that both of these channel types were dependent on the expression of {alpha}-ENaC, whereas the highly selective channels required the additional expression of β- and {gamma}-ENaC (26, 27, 32).

In the present study, however, single cells invariably displayed a nonselective conductance, whereas cells that had become integrated into small groups displayed an ENaC-like GNa (see also Ref. 14). Moreover, irrespective of the extent of cell-cell coupling, GNa was negligible in hormone-deprived cells (see also Ref. 14), and so our data indicate that the Na+-selective and -nonselective conductances are both induced by glucocorticoid stimulation/SGK1-S422D expression and subject to further upregulation by insulin/PI3K. These conductances therefore display parallel patterns of hormonal regulation, and our data, in contrast to the earlier studies (27, 32), thus suggest that hormonal stimulation has no influence over Na+ selectivity or amiloride sensitivity. Indeed, since all other culture conditions were identical, the present study indicates that electrical coupling between individual cells is the only factor that determines Na+ selectivity and amiloride sensitivity. This observation, when take together with the results of earlier studies (26, 27, 32), raises the possibility that contact between individual cells might be able to influence the relative abundance of the ENaC subunits in the plasma membrane, and there are a number of precedents for such independent control over the surface expression of {alpha}-, β-, and {gamma}-ENaC. For example, in alveolar type II cells, the β2-adrenoceptor agonist terbutaline increases in the surface abundance of β- and {gamma}-ENaC with no effect on {alpha}-ENaC (41). Moreover, although changes in O2 tension affect the surface expression of {alpha}-, β-, and {gamma}-ENaC, the effect on the {alpha}-subunit is smaller than the effect on β- and {gamma}-ENaC (41). Furthermore, infection with Mycoplasma pulmonis has been shown to suppress the amiloride-sensitive current that can be recorded from such cells, and this inhibitory response occurs with no change to the overall abundance of the {alpha}-, β-, and {gamma}-ENaC proteins. However, assays of surface expression showed that infection with M. pulmonis caused a fall in the surface abundance of {gamma}-ENaC that occurred with no effect on {alpha}-ENaC (21). Subsequent studies of H441 cells must therefore explore the effects of hormonal stimulation, SGK1/PI3K expression, and cell-cell contact on the surface expression of {alpha}-, β-, and {gamma}-ENaC.

Significance of present study. There are close parallels between the data presented here and the data presented by Shlyonsky et al. (48) who recorded currents from H441 cells grown to confluence in the presence of dexamethasone. H441 cells form "domes" under these conditions, and these structures appear as liquid absorbed from the overlying culture medium accumulates underneath the confluent cell sheet, lifting it clear of the culture substrate. Shlyonsky et al. (48) showed that the amiloride-sensitive, selective GNa could only be detected in cells that formed these domes, whereas cells located away from these structures expressed the La3+-sensitive, nonselective cation conductance. Moreover, these authors (48) showed that the dome-forming H441 cells express tight junctions and other markers of a polarized epithelial phenotype and therefore proposed that cell-cell communication was crucial to the development of the Na+-selective, amiloride-sensitive phenotype. Although both studies (the present study and Ref. 48) suggest contact between cells is vital for the formation of amiloride-sensitive, Na+-selective conductance, there are differences between the two sets of experiments. Most importantly, our data are derived entirely from subconfluent cells, and so we did not record any currents from cells in the dome-like structures described by Shlyonsky et al. (48). It therefore appears that H441 cells can be more readily induced to express the highly selective GNa in our hands, and this may well be a consequence of small differences in culture conditions. Interestingly, studies of mammary epithelial cells indicate that the process of dome formation is dependent on increased expression of β-ENaC. Normally, the expression of this gene appears to be repressed by a mechanism dependent on the product of a second gene, called 133, which encodes a separate, transmembrane protein (58), and dome formation appears to occur when 133 expression is repressed, which allows unimpeded β-ENaC expression (58). The way in which this process relates to the effects of cell contact described in the electrophysiological studies (the present study and Ref. 48) here is uncertain.

The present data confirm that glucocorticoids induce GNa in hormone-deprived H441 cells (14) and extend on these findings by showing that this effect is mimicked by heterologous expression of active SGK1 and by demonstrating that this glucocorticoid-induced conductance is subject to further regulation via insulin/PI3K. However, the present data also show that electrical coupling between cells exerts a powerful influence over the biophysical properties of this hormonally induced GNa, and such cell-cell interaction thus appears essential for the expression of the Na+-selective, amiloride-sensitive channels that permit Na+ entry in absorptive tissues.


   GRANTS
TOP
 ABSTRACT
 METHODS
 RESULTS
 DISCUSSION
 GRANTS
REFERENCES
 
We are grateful to the Wellcome Trust and Tenovus Scotland for the financial support that made this study possible. M. Gallacher thanks the George John and Sheilah Livanos Charitable Trust for a Prize PhD studentship.


   ACKNOWLEDGMENTS
 
We are grateful to Philip Cohen and Doreen Cantrell (College of Life Sciences, University of Dundee) for the provision of cDNA constructs and for their valuable advice concerning their use and to Sarah Inglis for helpful comments and suggestions.


   FOOTNOTES
 

Address for reprint requests and other correspondence: S. M. Wilson, Lung Membrane Transport Group, Division of Maternal and Child Health Sciences, Ninewells Hospital and Medical School, Univ. of Dundee, Dundee DD1 9SY, Scotland, United Kingdom (e-mail: S.M.Wilson{at}dundee.ac.uk)

The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.


   REFERENCES
TOP
 ABSTRACT
 METHODS
 RESULTS
 DISCUSSION
 GRANTS
 REFERENCES
 

  1. Alvarez De La Rosa D, Zhang P, Náray-Fejes-Tóth A, Fejes-Tóth G, Canessa CM. The serum and glucocorticoid kinase sgk increases the abundance of epithelial sodium channels in the plasma membrane of Xenopus oocytes. J Biol Chem 274: 37834–37839, 1999.[Abstract/Free Full Text]
  2. Alvarez De La Rosa D, Canessa CM. Role of SGK in hormonal regulation of epithelial sodium channel in A6 cells. Am J Physiol Cell Physiol 284: C404–C414, 2003.[Abstract/Free Full Text]
  3. Alvarez De La Rosa D, Paunescu TG, Els WJ, Helman SI, Canessa C. Mechanisms or regulation of epithelial sodium channel by SGK1 in A6 cells. J Gen Physiol 124: 395–407, 2004.[Abstract/Free Full Text]
  4. Arteaga MF, Canessa C. Functional specificity of Sgk1 and Akt1 on ENaC activity. Am J Physiol Renal Physiol 289: F90–F96, 2005.[Abstract/Free Full Text]
  5. Barker PM, Brown MJ, Ramsden CA, Strang LB, Walters DV. Synergistic action of triiodothyronine and hydrocortisone on epinephrine-induced reabsorption of fetal lung liquid. Pediatr Res 27: 588–591, 1990.[Web of Science][Medline]
  6. Barry PH, Lynch JW. Liquid junction potentials and small cell effects in patch-clamp analysis. J Membr Biol 121: 101–117, 1991.[CrossRef][Web of Science][Medline]
  7. Blazer-Yost BL, Liu X, Helman SI. Hormonal regulation of ENaCs: insulin and aldosterone. Am J Physiol Cell Physiol 274: C1373–C1379, 1998.[Abstract/Free Full Text]
  8. Blazer-Yost BL, Esterman MA, Vlahos CJ. Insulin-stimulated trafficking of ENaC in renal cells requires PI 3-kinase activity. Am J Physiol Cell Physiol 284: C1645–C1653, 2003.[Abstract/Free Full Text]
  9. Boucher RC. Molecular insights into the physiology of the 'thin film' of airway surface liquid. J Physiol 516: 631–638, 1999.[Abstract/Free Full Text]
  10. Brown MJ, Olver RE, Ramsden CA, Strang LB, Walters DV. Effects of adrenaline and of spontaneous labour on the secretion and absorption of lung liquid in the fetal lamb. J Physiol 344: 137–152, 1983.[Abstract/Free Full Text]
  11. Chambers LA, Constable MJ, Clunes MT, Ko WH, Olver RE, Inglis SK, Wilson SM. Adenosine-evoked Na+ transport in airway epithelial cells. Br J Pharmacol 149: 43–55, 2006.[CrossRef][Web of Science][Medline]
  12. Chambers LA, Rollins BM, Tarran R. Liquid movement across the surface epithelium of large airways. Respir Physiol Neurobiol 159: 256–270, 2007.[CrossRef][Web of Science][Medline]
  13. Cheng G, Kim MJ, Jia G, Agrawal DK. Involvement of chloride channels in IGF-I-induced proliferation of porcine arterial smooth muscle cells. Cardiovasc Res 73: 198–207, 2007.[Abstract/Free Full Text]
  14. Clunes MT, Butt AG, Wilson SM. A glucocorticoid-induced Na+ conductance in human airway epithelial cells identified by perforated patch recording. J Physiol 557: 809–819, 2004.[Abstract/Free Full Text]
  15. Cohen P. The twentieth century struggle to decipher insulin signalling. Nat Rev Mol Cell Biol 7: 867–873, 2006.[CrossRef][Web of Science][Medline]
  16. Di Stefano A, Wittner M, Schlatter E, Lang HJ, Englert H, Greger R. Diphenylamine-2-carboxylate, a blocker of the Cl–-conductive pathway in Cl–-transporting epithelia. Pflügers Arch 405, Suppl 1: S95–S100, 1985.[CrossRef][Web of Science][Medline]
  17. Guazzi M, Bramilla R, De Vita S, Guazzi MD. Diabetes worsens pulmonary diffusion in heart failure, and insulin counteracts this effect. Am J Respir Crit Care Med 166: 978–982, 2002.[Abstract/Free Full Text]
  18. Guazzi MD, Oreglia I, Guazzi MD. Insulin improves alveolar-capillary membrane gas conductance in type 2 diabetes. Diabetes Care 25: 1802–1806, 2002.[Abstract/Free Full Text]
  19. Hagiwara N, Tohda H, Doi Y, O'Brodovich H, Marunaka Y. Effects of insulin and tyrosine kinase inhibitor on ion transport in the alveolar cell of the fetal lung. Biochem Biophys Res Commun 187: 302–308, 1992.
  20. Helms MN, Liu L, Liang YY, Al-Khalili O, Vandewalle A, Saxena S, Eaton DC, Ma HP. Phosphatidylinositol 3,4,5-trisphosphate mediates aldosterone stimulation of epithelial sodium channel (ENaC) and interacts with {gamma}-ENaC. J Biol Chem 280: 40885–40891, 2005.[Abstract/Free Full Text]
  21. Hickman-Davis JM, McNicholas-Bevensee CM, Davis IC, Ma HP, Davis GC, Bosworth CA, Matalon S. Reactive species mediate inhibition of alveolar type II sodium transport during mycoplasma infection. Am J Respir Crit Care Med 173: 334–344, 2006.[Abstract/Free Full Text]
  22. Hilgemann DW. Getting ready for the decade of lipids. Annu Rev Physiol 65: 697–700, 2003.[CrossRef][Web of Science][Medline]
  23. Horn R, Marty A. Muscarinic activation of ionic currents measured by a new whole-cell recording method. J Gen Physiol 92: 145–159, 1988.[Abstract/Free Full Text]
  24. Inglis SK, Brown SG, Constable MJ, McTavish N, Olver RE, Wilson SM. A Ba2+-resistant, acid-sensitive K+ conductance in Na+-absorbing H441 human airway epithelial cells. Am J Physiol Lung Cell Mol Physiol 292: L1304–L1312, 2007.[Abstract/Free Full Text]
  25. Ishikawa T, Marunaka Y, Rotin D. Electrophysiological characterization of the rat epithelial Na+ channel (rENaC) expressed in MDCK cells. Effects of Na+ and Ca2+. J Gen Physiol 111: 825–846, 1998.[Abstract/Free Full Text]
  26. Jain L, Chen XC, Malik B, Al-Khalili O, Eaton DC. Antisense oligonucleotides against the {alpha}-subunit of ENaC decrease lung epithelial cation channel activity. Am J Physiol Lung Cell Mol Physiol 276: L1046–L1051, 1999.[Abstract/Free Full Text]
  27. Jain L, Chen XL, Ramosevac S, Brown LA, Eaton DC. Expression of highly selective sodium channels in alveolar type II cells is determined by culture conditions. Am J Physiol Lung Cell Mol Physiol 280: L646–L658, 2001.[Abstract/Free Full Text]
  28. Kilic G, Doctor RB, Fitz JG. Insulin stimulates membrane conductance in a liver cell line. Evidence for insertion of ion channels through a phosphoinositidide 3-kinase-dependent mechanism. J Biol Chem 276: 26762–26768, 2001.[Abstract/Free Full Text]
  29. Kobayashi T, Cohen P. Activation of serum- and glucocorticoid-regulated protein kinases by agonists that activate phosphatidylinositol 3-kinase is mediated by 3-phosphoinositide-dependent protein kinase 1 (PDK1) and PDK2. Biochem J 339: 319–328, 1999.[CrossRef][Web of Science][Medline]
  30. Kulaksiz H, Schmid A, Hönscheid M, Ramaswamy A, Cetin Y. Clara cell impact in air side activation of CFTR in small pulmonary airways. Proc Natl Acad Sci USA 99: 6796–6801, 2002.[Abstract/Free Full Text]
  31. Lang F, Cohen P. Regulation and physiological roles of serum- and glucocorticoid-induced protein kinase isoforms. Sci STKE 108: RE17, 2001.
  32. Lazrak A, Samanta A, Venetsanou K, Barbary P, Matalon S. Modification of biophysical properties of lung epithelial Na+ channels by dexamethasone. Am J Physiol Cell Physiol 279: C762–C770, 2000.[Abstract/Free Full Text]
  33. Lazrak A, Matalon S. cAMP-induced changes of apical membrane potentials of confluent H441 monolayers. Am J Physiol Lung Cell Mol Physiol 285: L443–L450, 2003.[Abstract/Free Full Text]
  34. Mall M, Grubb BR, Harkema JR, O'Neal WK, Boucher RC. Increased airway epithelial Na+ absorption produces cystic fibrosis-like lung disease in mice. Nat Med 10: 487–493, 2004.[CrossRef][Web of Science][Medline]
  35. Matthay MA, Wiener-Kronish JP. Intact epithelial barrier function is critical for the resolution of pulmonary edema in man. Am Rev Respir Dis 142: 1250–1257, 1990.[Web of Science][Medline]
  36. Muanprasat C, Sonawane ND, Salinas D, Taddei A, Galietta LJV, Verkman AS. Discovery of glycine hydrazide pore-occluding CFTR inhibitors: mechanism, structure-activity analysis, and in vivo efficacy. J Gen Physiol 124: 125–137, 2004.[Abstract/Free Full Text]
  37. Nofziger C, Chen L, Shane MA, Smith CD, Brown KK, Blazer-Yost BL. PPAR{gamma} agonists do not directly enhance basal or insulin-stimulated Na+ transport via the epithelial Na+ channel. Pflügers Arch 451: 445–453, 2005.[CrossRef][Web of Science][Medline]
  38. O'Brodovich H. Immature epithelial Na+ channel expression is one of the pathological mechanisms leading to human neonatal respiratory distress syndrome. Proc Assoc Am Physicians 108: 345–355, 1996.[Web of Science][Medline]
  39. Olver RE, Walters DV, Wilson SM. Developmental regulation of lung liquid transport. Annu Rev Physiol 66: 77–101, 2004.[CrossRef][Web of Science][Medline]
  40. Paunescu TG, Blazer-Yost BL, Vlahos CJ, Helman SI. LY-294002-inhibitable PI 3-kinase and regulation of baseline rates of Na+ transport in A6 epithelia. Am J Physiol Cell Physiol 279: C236–C247, 2000.[Abstract/Free Full Text]
  41. Planès C, Blot-Chabaud M, Matthay MA, Couette S, Uchida T, Clerici C. Hypoxia and beta2-agonists regulate cell surface expression of the epithelial sodium channel in native alveolar epithelial cells. J Biol Chem 277: 47318–47324, 2002.[Abstract/Free Full Text]
  42. Pochynyuk O, Staruschenko A, Tong Q, Medina J, Stockand JD. Indentification of a functional phosphatidylinositol 3,4,5-trisphosphate binding site in the epithelial Na+ channel. J Biol Chem 280: 37565–37571, 2005.[Abstract/Free Full Text]
  43. Pochynyuk O, Tong Q, Staruschenko A, Stockand JD. Binding and direct activation of the epithelial Na+ channel (ENaC) by phosphatidylinositides. J Physiol 580: 365–372, 2007.[Abstract/Free Full Text]
  44. Qu ZQ, Hartzell HC. Functional geometry of the permeation pathway of Ca2+-activated Cl– channels inferred from analysis of voltage-dependent block. J Biol Chem 276: 18423–18429, 2001.[Abstract/Free Full Text]
  45. Ramminger SJ, Richard K, Inglis SK, Land SC, Olver RE, Wilson SM. A regulated apical Na+ conductance in dexamethasone-treated H441 airway epithelial cells. Am J Physiol Lung Cell Mol Physiol 287: L411–L419, 2004.[Abstract/Free Full Text]
  46. Record RD, Froelich LL, Vlahos CJ, Blazer-Yost BL. Phosphatidylinositol 3-kinase activation is required for insulin-stimulated sodium transport in A6 cells. Am J Physiol Endocrinol Metab 274: E611–E617, 1998.[Abstract/Free Full Text]
  47. Reif K, Nobes CD, Thomas G, Hall A, Cantrell DA. Phosphatidylinositol 3-kinsase signals activate a selective subset of Rac/Rho-dependent effector pathways. Curr Biol 6: 1445–1455, 1996.[CrossRef][Web of Science][Medline]
  48. Shlyonsky V, Goolaerts A, Van Beneden R, Sariban-Sohraby S. Differentiation of epithelial Na+ channel function: an in vitro model. J Biol Chem 280: 24181–24187, 2005.[Abstract/Free Full Text]
  49. Snyder PM, Olson DR, Kabra R, Zhou R, Steines JC. cAMP and serum and glucocorticoid-inducible kinase (SGK) regulate the epithelial Na+ channel through convergent phosphorylation of Nedd4–2. J Biol Chem 279: 45753–45758, 2004.[Abstract/Free Full Text]
  50. Snyder PM, Steines JC, Olson DR. Relative contribution of Nedd4 and Nedd4–2 to ENaC regulation in epithelia determined by RNA interference. J Biol Chem 279: 5042–5046, 2004.[Abstract/Free Full Text]
  51. Thomas CP, Campbell JR, Wright PJ, Husted RF. cAMP-stimulated Na+ transport in H441 distal lung epithelial cells: role of PKA, phosphatidylinositol 3-kinase, and sgk1. Am J Physiol Lung Cell Mol Physiol 287: L843–L851, 2004.[Abstract/Free Full Text]
  52. Tong Q, Gamper N, Medina JL, Shapiro MS, Stockand JD. Direct activation of the epithelial Na+ channel by phosphatidylinositol 3,4,5-trisphosphate and phoshatidylinositol3,4-bisphosphate produced by phosphoinositide 3-OH kinase. J Biol Chem 279: 22654–22663, 2004.[Abstract/Free Full Text]
  53. Wang J, Barbary P, Maiyar A, Rozansky DJ, Bhargava A, Leong M, Firestone GL, Pearce D. SGK integrates insulin and mineralocorticoid regulation of epithelial sodium transport. Am J Physiol Renal Physiol 280: F303–F313, 2001.[Abstract/Free Full Text]
  54. Ware LB, Matthay MA. Alveolar epithelial fluid clearance is impaired in the majority of patients with acute lung injury and acute respiratory distress syndrome. Am J Respir Crit Care Med 163: 1376–1383, 2001.[Abstract/Free Full Text]
  55. Webb KF, Merriman-Smith BR, Stobie JK, Kistler J, Donaldson PJ. Cl– influx into rat cortical lens fiber cells is mediated by a Cl– conductance that is not ClC-2 or -3. Invest Ophthalmol Vis Sci 45: 4400–4408, 2004.[Abstract/Free Full Text]
  56. Wilson SM, Brown SG, McTavish N, McNeill RP, Husband EM, Inglis SK, Olver RE, Clunes MT. Expression of the intermediate conductance, Ca2+-activated K+ channel (KCNN4) in H441 human distal airway epithelial cells. Am J Physiol Lung Cell Mol Physiol 291: L957–L965, 2006.[Abstract/Free Full Text]
  57. Wulff P, Vallon V, Huang DY, Volkl H, Yu F, Richter K, Jansen M, Schlunz M, Kauselmann G, Bosl MR, Lang F, Kuhl D. Impaired renal Na+ retention in the sgk1-knockout mouse. J Clin Invest 110: 1263–1268, 2002.[CrossRef][Web of Science][Medline]
  58. Zucchi I, Montagna C, Susani L, Montesano R, Affer M, Zanotti S, Redolfi E, Vezzoni P, Dulbecco R. Genetic dissection of dome formation in a mammary cell line: identication of two genes with opposing action. Proc Natl Acad Sci USA 96: 13766–13770, 1999.[Abstract/Free Full Text]



This article has been cited by other articles:


Home page
 J. Physiol.Home page
M. Gallacher, S. G. Brown, B. G. Hale, R. Fearns, R. E. Olver, R. E. Randall, and S. M. Wilson
Cation currents in human airway epithelial cells induced by infection with influenza A virus
J. Physiol., July 1, 2009; 587(13): 3159 - 3173.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
294/5/L942    most recent
00240.2007v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Web of Science (2)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Brown, S. G.
Right arrow Articles by Wilson, S. M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Brown, S. G.
Right arrow Articles by Wilson, S. M.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Visit Other APS Journals Online
Copyright © 2008 by the American Physiological Society.